Language Model Crossover: Variation through Few-Shot Prompting

Abstract

This paper pursues the insight that language models naturally enable an intelligent variation operator similar in spirit to evolutionary crossover. In particular, language models of sufficient scale demonstrate in-context learning, i.e. they can learn from associations between a small number of input patterns to generate outputs incorporating such associations (also called few-shot prompting). This ability can be leveraged to form a simple but powerful variation operator, i.e. to prompt a language model with a few text-based genotypes (such as code, plain-text sentences, or equations), and to parse its corresponding output as those genotypes' offspring. The promise of such language model crossover (which is simple to implement and can leverage many different open-source language models) is that it enables a simple mechanism to evolve semantically-rich text representations (with few domain-specific tweaks), and naturally benefits from current progress in language models. Experiments in this paper highlight the versatility of language-model crossover, through evolving binary bit-strings, sentences, equations, text-to-image prompts, and Python code. The conclusion is that language model crossover is a promising method for evolving genomes representable as text.

References in corpus (15)

• Scaling Instruction-Finetuned Language Models
• CoCa: Contrastive Captioners are Image-Text Foundation Models
• Symbolic Regression in Materials Science
• Image Aesthetic Assessment: An Experimental Survey
• Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answering
• Bayesian Optimization is Superior to Random Search for Machine Learning Hyperparameter Tuning: Analysis of the Black-Box Optimization Challenge 2020
• Where are we now? A large benchmark study of recent symbolic regression methods
• Language Model Crossover: Variation through Few-Shot Prompting
• LLMatic: Neural Architecture Search via Large Language Models and Quality Diversity Optimization
• Towards Automated Circuit Discovery for Mechanistic Interpretability
• Differentiable Quality Diversity
• Policy Manifold Search: Exploring the Manifold Hypothesis for Diversity-based Neuroevolution
• MarioGPT: Open-Ended Text2Level Generation through Large Language Models
• EvoPrompting: Language Models for Code-Level Neural Architecture Search
• Simple Genetic Operators are Universal Approximators of Probability Distributions (and other Advantages of Expressive Encodings)

Cited by in corpus (3)

• Language Model Crossover: Variation through Few-Shot Prompting
• Can Large Language Models Be Trusted as Evolutionary Optimizers for Network-Structured Combinatorial Problems?
• LLM2TEA: An Agentic AI Designer for Discovery with Generative Evolutionary Multitasking